Dynamics of the formation of a charge transfer state in 1,2-bis(9-anthryl)acetylene in polar solvents: Symmetry reduction with the participation of an intramolecular torsional coordinate

Article abstract of DOI:10.1021/jp4038705

We have studied 1,2-bis(9-anthryl)acetylene as a model compound for the characterization of the process of solvent-mediated symmetry reduction in an excited state. Thanks to the acetylenic bridge that joins the two anthracenic moieties, this system maintains minimal steric hindrance between the end chromophores in comparison with the classic 9,9a?2-bianthryl model compound. The acetylenic bridge also allows for significant electronic coupling across the molecule, which permits a redistribution of electron density after light absorption. Femtosecond resolved fluorescence measurements were used to determine the spectral evolution in acetonitrile and cyclohexane solutions. We observed that, for 1,2-bis(9-anthryl)acetylene, the formation of a charge transfer state occurs in a clear bimodal fashion with well separated time scales. Specifically, the evolution of the emission spectrum involves a first solvent-response mediated subpicosecond stage where the fluorescence changes from that typical of nonpolar solvents (locally excited) to an intermediate, partial charge transfer state. The second stage of the evolution into a full charge transfer state occurs with a much longer time constant of 37.3 ps. Since in this system the steric hindrance is minimized, this molecule can undergo much larger amplitude motions for the torsion between the two anthracenic moieties associated with the charge redistribution in comparison with the typical model compound 9,9a?2-bianthryl. Clearly, the larger range of motions of 1,2-bis(9-anthryl)acetylene gives the opportunity to study the electron transfer process with a good separation of the time scales for the formation of a partial charge transfer state, determined by the speed of solvent response, and the intramolecular changes associated with the formation of the fully equilibrated charge transfer state. ? 2013 American Chemical Society.

Full text of DOI:10.1021/jp4038705

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Article
pubs.acs.org/JPCB
Dynamics of the Formation of a Charge Transfer State in 1,2-Bis(9-
anthryl)acetylene in Polar Solvents: Symmetry Reduction with the
Participation of an Intramolecular Torsional Coordinate
Luis Gutierrez-Arzaluz,† Cesar A. Guarin,† William Rodríguez-Cordoba,†,‡ and Jorge Peon*,†
́
́
†Instituto de Química, Universidad Nacional Auton
Mexico
́ ́ ́
oma de Mexico, Circuito Exterior, Ciudad Universitaria, Mexico, 04510, D.F.,
́
‡Escuela de Física, Universidad Nacional de Colombia, Sede Medellín, A.A. 3840, Medellín, Colombia
S
* Supporting Information
ABSTRACT: We have studied 1,2-bis(9-anthryl)acetylene as
a model compound for the characterization of the process of
solvent-mediated symmetry reduction in an excited state.
Thanks to the acetylenic bridge that joins the two anthracenic
moieties, this system maintains minimal steric hindrance
between the end chromophores in comparison with the classic
9,9′-bianthryl model compound. The acetylenic bridge also
allows for significant electronic coupling across the molecule,
which permits a redistribution of electron density after light
absorption. Femtosecond resolved fluorescence measurements
were used to determine the spectral evolution in acetonitrile
and cyclohexane solutions. We observed that, for 1,2-bis(9-
anthryl)acetylene, the formation of a charge transfer state
occurs in a clear bimodal fashion with well separated time scales. Specifically, the evolution of the emission spectrum involves a
first solvent-response mediated subpicosecond stage where the fluorescence changes from that typical of nonpolar solvents
(locally excited) to an intermediate, partial charge transfer state. The second stage of the evolution into a full charge transfer state
occurs with a much longer time constant of 37.3 ps. Since in this system the steric hindrance is minimized, this molecule can
undergo much larger amplitude motions for the torsion between the two anthracenic moieties associated with the charge
redistribution in comparison with the typical model compound 9,9′-bianthryl. Clearly, the larger range of motions of 1,2-bis(9-
anthryl)acetylene gives the opportunity to study the electron transfer process with a good separation of the time scales for the
formation of a partial charge transfer state, determined by the speed of solvent response, and the intramolecular changes
associated with the formation of the fully equilibrated charge transfer state.
separated state where the solvent shells stabilize positive and
negative charges localized respectively on the two anthracenic
rings.9−11 This results in a bathochromic fluorescence shift due
to the stabilization of the dipolar emissive state from the
interaction with the polar solvent molecules.
The first time-resolved fluorescence experiments on BA
suggested that the dynamics of the electron transfer reaction are
governed mainly by solvation.7,8 More recent pump−probe
studies have indicated that the excited state evolution occurs in
two different stages, namely, locally excited (LE, nonpolar) →
intermediate or partial charge transfer state (PCT) → fully
relaxed charge transfer state (CT). The corresponding kinetic
models for this process can include the possibility of the initial
setup of an equilibrium between the LE and the intermediate
PCT state.5,12−14 Such representation is based on BA’s
INTRODUCTION
■
Light induced separation of charges is one of the most
fundamental processes in photochemistry and plays a central
role in photosynthetic and photovoltaic applications.1−3 In the
search for a fundamental understanding of this process, model
symmetric systems like 9,9′-bianthryl (BA, see Scheme 1) have
played a central role.4−13 In this kind of molecule, an initially
formed electronically excited state (highly polarizable locally
excited, nonpolar) can evolve toward a highly polar, charge
Scheme 1. Structures of (a) 9,9′-Bianthryl (BA) and (b) 1,2-
Bis(9-anthryl)acetylene (BisAA)
Received: April 18, 2013
Revised: September 6, 2013
Published: September 9, 2013
© 2013 American Chemical Society
12175
dx.doi.org/10.1021/jp4038705 | J. Phys. Chem. B 2013, 117, 12175−12183

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry B
Article
emission spectrum evolution,13 transient absorption,6,14 and
anisotropy measurements.12 More specifically, in the transient
absorption experiments by Kovalenko et al.,14 it was observed
that although the spectral evolution is consistent with a two
stage model (with a rapid initial equilibrium involving the LE
state), both the first rapid stage and the latter stage are solvent
controlled. That study assigned the sub 60 fs step in acetonitrile
to a high frequency solvent coordinate, and the latter stage
(subpicosecond in acetonitrile) to further solvent relaxation
(diffusive or dissipative motions of the acetonitrile solvent
shells). Importantly, the transient absorption experiments of
Kovalenko et al.14 on BA do not distinguish evidence for the
participation of an intramolecular torsional motion around the
anthracene−anthracene single bridging bond related to the
formation of the CT state. The use of BA as a model for charge
transfer processes, therefore, does not permit a clear study of
the role of intramolecular rearrangements which may be crucial
and more general in CT dynamics (see below).
Scheme 2. Synthesis of 1,2-Di(anthracen-9-yl)ethyne or 1,2-
Bis(9-anthryl)acetylene (5)
In this contribution we propose a new model compound for
the study of the photoinduced symmetry reduction in polar
solvents which can give much clearer information about the
involvement of an intramolecular torsional coordinate. This
compound, 1,2-bis(9-anthryl)acetylene (BisAA, see Scheme 1),
maintains the fundamental properties of BA of a non-dipolar
ground state due to the nuclear symmetry, and extended
conjugation which allows for the charge transfer process. The
BisAA molecule, however, has a much lower degree of
restriction about its geometry in comparison with BA, where,
due to the steric hindrance between the anthracenic units in BA
(from the peri-hydrogens and their proximity), the ground state
of this molecule has a perpendicular disposition between the
two anthracenes (D2d symmetry). In fact, from the hindrance,
the electronically excited state of BA can evolve only within a
minimal margin of conformations which by far cannot approach
planar or near-planar geometries. This steric restriction in BA
implies than any intramolecular geometric change is of minimal
amplitude and therefore may occur (if at all) in time scales
where the fast solvent motions are also occurring.
As mentioned, BisAA has a much more free range of motion
for the inter-anthracene torsional coordinate, and also includes
in the bridge an acetylenic triple bond which can accommodate
these rotations.15 As we show through femtosecond fluo-
rescence upconversion experiments, in this molecule we
observed two well separated time scales for the formation of
the CT state in acetonitrile. Since the slower stage is observed
to occur in a widely different time scale from the fast one (most
likely related to the solvent response), this system allows for the
direct observation of the participation of an intramolecular
torsional coordinate that can be included in advanced models
for the charge transfer process.
refluxing toluene. Finally, the resulting terminal acetylene (4)
was coupled with 9-bromoanthracene formed in the same
coupling procedure used before to obtain (3), to yield the
desired 1,2-di(anthracen-9-yl)ethyne or 1,2-bis(9-anthryl)-
acetylene, BisAA (5),21−23 in 68% in the last reaction step.
All physical properties and spectroscopic data agree with the
values reported in the literature. More details about the
synthetic procedures and characterization for each step are
included in the Supporting Information.
Steady-state absorption spectra were acquired in a Cary 50
spectrometer, and fluorescence spectra were acquired in a Cary-
Eclipse fluorimeter (Varian). Our fluorescence upconversion
setup has been described in detail elsewhere.24−26 For
excitation we used the second harmonic pulses from a
regeneratively amplified Ti:sapphire laser with a 1 kHz
repetition rate. The polarization of the pump-pulse was
adjusted to magic angle conditions with respect to the
detection axis (vertical) with a 400 nm half-wave plate, and
then focused to the sample with a 30 cm focal length lens. The
solutions were studied in a rapid flow quartz cell (2 mL/s) of 1
mm optical path length. The fluorescence from the sample was
collected with a pair of parabolic mirrors and focused in a 0.5
mm β-BBO upconversion crystal where it was gated with pulses
separated from the fundamental beam (800 nm) and time
delayed. The polarization of the gate pulses was adjusted to the
vertical direction (crystal’s ordinary axis) with a half-wave plate.
The resulting sum-frequency mixing signal was collected with a
5 cm diameter lens and focused into a double monochromator
to be detected with a photomultiplier tube. The detector signal
was passed through an oscilloscope and connected to a lock-in
amplifier (Stanford Research Systems). The signal was
EXPERIMENTAL SECTION
■
Methods and Materials. 1,2-Di(anthracen-9-yl)ethyne,
also named 1,2-bis(9-anthryl)acetylene (BisAA (5)), was
synthesized using the four steps presented in Scheme 2.
Briefly, anthracene (1) was converted to 9-bromoanthracene
(2) in the presence of aqueous hydrohalic acid and hydrogen
peroxide following the bromination methodology reported by
Vyas et al.16 Subsequently, reaction of 9-bromoanthracene with
2-methylbut-3-yn-2-ol in a standard Sonogashira coupling
produced the corresponding 4-(anthracen-10-yl)-2-methylbut-
3-yn-2-ol (3),9,17−20 which was converted to the 9-ethynylan-
thracene (4)18,19 by treatment with potassium hydroxide in
12176
dx.doi.org/10.1021/jp4038705 | J. Phys. Chem. B 2013, 117, 12175−12183

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry B
Article
modulated to 1/3 of the laser repetition rate by a phase locked
optical chopper (New Focus). The upconversion instrumental
response function is mainly limited by the optical path length in
the cell and was determined from upconversion of the solvent
Raman signals to be Gaussian with a full width at half-maximum
of 490 fs. In all cases, back-to-back scans with the solvent only
were acquired to ensure that the signals are not contaminated
by the upconversion of scattered light or other artifacts.
Computational Method. The ground state geometry of
BisAA was optimized at the B3LYP/6-311+G(d,p) level of
theory. The effects of the solvent were taken into account
through the polarizable continuum model IEFPCM. The
energies of different conformations with a varying dihedral
angle between the anthracenic moieties were calculated in the
electronic ground state. For this, the twist angle was set
constant at different values and the rest of the degrees of
freedom were optimized. The Gaussian 09 set of programs was
used for all calculations.27 It should be noticed that we only
include calculations for the ground state in order to understand
some details of its potential energy surface. From previous
studies in BA, the respective calculations of the excited states of
these symmetric model compounds in solution will require the
use of highly specialized methods which include explicit solvent
molecules and account for their polarizability.28
low energy band at 441 nm and an intense band at 266 nm.
Such similarity with the anthracene absorption spectrum is
consistent with an anthracene-localized transition, similar to the
case of BA.28 The red shifting of the first absorption of BisAA in
comparison with anthracene (356 nm) can be explained as an
effect of a substitution on the short molecular axis of this
chromophore which is the direction of the respective transition
dipole moment. A localized transition is expected if the BisAA
molecule does not retain a fully planar structure as observed in
the crystal phase.21,29 Theoretical calculations at the density
functional level of theory (see Figure 3) indicate that, in fact, in
Figure 3. Potential energy of electronic ground state for BisAA as
function of the twist angle between the anthryl groups in the gas
phase, in cyclohexane and acetonitrile solutions. The level of theory
was B3LYP/(IEFPCM)/6-311+G(d,p).
RESULTS AND DISCUSSION
■
Figures 1 and 2 show the absorption and emission spectra of
BisAA in cyclohexane and acetonitrile. As can be seen, the
the ground state, BisAA does not correspond to a fully planar
conformation in gas and solution phases, but that, instead, an
angle between the two anthryl units of about 38° exists (38.2°
in the gas phase, 38.0° in cyclohexane, 37.3° in acetonitrile).
Small energy barriers for BisAA at this level of theory are
observed for torsional angles of 0° and near 90°. These barriers
are approximately 0.4 and 1.2 kcal/mol respectively in the gas
phase. As has been explained previously,15 such small barriers
imply a preferential conformation in acetylene-bridged systems
(at near 38° BisAA), although they are small enough to
produce a distribution of conformations.
Differently from the absorption spectra, for the emission
spectra, clear red shifting and shape changes take place when
going from cyclohexane to acetonitrile solutions. Particularly,
the fluorescence spectrum in cyclohexane maintains clear
anthracenic type vibronic structure, while the acetonitrile
solutions show a broad emission which is Stokes shifted by
1822 cm−1, corresponding to a CT state type emission. This
emission is very similar to the one for BA in polar solvents.5,6
We have performed studies of the emission in cyclohexane and
acetonitrile as a function of concentration to verify the absence
of effects like ground state association or the formation of
excimers. The results are included in the Supporting
Information where the concentration of BisAA was varied by
more than 7 orders of magnitude below the working
concentration of the upconversion measurements. The
emission spectra do not show any changes as a function of
concentration, which is indicative of the absence of
intermolecular processes like the formation of aggregates or
excimers.
Figure 1. Absorption spectra of BisAA in cyclohexane and acetonitrile
solutions. The inset shows details of the first absorption band.
Figure 2. Emission of BisAA in cyclohexane and acetonitrile solutions.
The excitation wavelength was 400 nm.
absorption spectra are not significantly altered in going from a
nonpolar to a polar solvent. This is expected from a symmetric
non-dipolar electronic ground state. The absorption spectrum
of BisAA is similar to that of anthracene with the presence of a
Figures 4 to 6 show the emission spectrum of BisAA in a
series of solvents. In solvents of low polarity like dichloro-
methane (DCM) and tetrahydrofuran (THF), the emission
spectra of BisAA appear to be formed by a superposition of the
12177
dx.doi.org/10.1021/jp4038705 | J. Phys. Chem. B 2013, 117, 12175−12183

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry B
Article
Further, we have employed the method of Kawski and co-
workers to estimate the dipole moments of the ground and CT
excited state of BisAA in polar solvents.30−35 For this, we
measured the spectral position of the first absorption and the
CT state type emission in a series of polar solvents (Figure 6).
The emission data (vabs − vfluor) and (vabs + vfluor) was plotted as
̅
̅
̅
̅
a function of the polarity parameters f(ε,n) (eq 1) and the sum
f(ε,n) + 2g(n) (eq 2) as shown in Figure 7,
2
2
⎡
⎤
2n + 1 ε − 1
n − 1
f(ε, n) =
−
⎢
⎥
⎦
n2 + 2
n + 2
2
⎣ ε + 2
(1)
(2)
Figure 4. Emission spectra of BisAA in tetrahydrofuran (black) and
calculated (red) from contributions in cyclohexane (0.22) and
acetonitrile (0.78). Inset: weighted area normalized emission spectra
of BisAA in cyclohexane, acetonitrile (ACN), and acetonitrile left
shifted by 10 nm (ACN shifted).
n4 − 1
⎣(n + 2) ⎦
⎡
⎢
⎤
3
g(n) =
2
⎥
2
2
Figure 7. Plots of (a) Stokes shift (va − vf) and (b) sum of absorption
̅
̅
and emission maxima (va + vf) as a function of the solvent polarity
̅
̅
Figure 5. Emission spectra of BisAA in dichloromethane (DCM,
black) and calculated (red) from contributions in cyclohexane (0.19)
and acetonitrile (0.81). Inset: weighted area normalized emission
spectra of BisAA in cyclohexane, acetonitrile (ACN), and acetonitrile
left shifted by 10 nm (ACN shifted).
parameters f(ε,n) and f(ε,n) + 2g(n) respectively. Solvents in (a) and
(b): (1) tetrahydrofuran, (2) dichloromethane, (3) isopropanol, (4)
methanol, (5) propanol, (6) ethanol, (7) acetone, (8) acetonitrile, (9)
N,N-dimethylformamide, and (10) dimethyl sulfoxide.
where n is the refraction index and ε the dielectric constant of
each solvent. The method of Kawski et al.35 takes the slopes
from the linear fits of both plots: m1 from Figure 7(a) and m2
from Figure 7b to estimate the dipolar moment in the ground
and CT emissive states using eqs 3 and 4 respectively:
hca3
2m1
m2 − m1
μ
=
ground
2
(3)
(m2 + m1)
m2 − m1
μ
= μ
excited
ground
(4)
where a is the radius of the respective Onsager cavity radius
(here taken as 5.95 Å). The result of this analysis estimates a S0
dipole moment of 1.1 0.90 D and of 9.7 2.9 D for the CT
state (the uncertainties were propagated from the uncertainties
in the slopes in the linear regression analysis). The
approximately 9.7 D dipole moment for the CT state in
BisAA is similar to that determined for BA of approximately 8
D.5,36 This analysis of the Stokes shifting of the CT band
confirms the highly dipolar nature of the emissive state in polar
solvents in this molecule.
Turning to the time-resolved studies, the femtosecond
fluorescence upconversion results for BisAA in cyclohexane
and acetonitrile solutions are shown in Figures 8 to 11. The
spectral evolutions of the emission bands were constructed
from the single wavelength upconversion scans together with
the corresponding normalization procedure which uses the
steady-state spectra and multiexponential fits to the single
Figure 6. Steady-state emission spectra for BisAA solutions in several
polar solvents. DMF: N,N-dimethylformamide.
cyclohexane and acetonitrile type emissions. In fact, as shown in
Figures 4 and 5, the fluorescence from the THF and DCM
solutions can be approximated by a weighted sum of the
emission in cyclohexane, plus the emission in acetonitrile after a
10 nm shift to shorter wavelengths. The necessity of shifting
the acetonitrile component to produce a good match between
the constructed spectrum (cyclohexane emission + shifted
acetonitrile emission) with the THF and DCM spectra is due to
the dependence of the CT type emission on the solvent polarity
(see below). The observation that, in these low polarity
solvents, both LE and unstructured red-shifted type emissions
are observed is indicative of an equilibrium between two types
of states.
12178
dx.doi.org/10.1021/jp4038705 | J. Phys. Chem. B 2013, 117, 12175−12183